modeling of engineered systems inthe vadosezone
TRANSCRIPT
SRNL-MS-2010-00070
Modeling of Engineered Systems in the Vadose Zone
Greg Flach
13 April 2010
Richland WA
Performance Assessment Community of Practice Technical Exchange
2
SRNL-MS-2010-00070
Outline
Engineered Systems at the Savannah River Site
Key failure / degradation modes
Modeling philosophy
Modeling practice
Opportunities for ASCEM and CBP
3
SRNL-MS-2010-00070
Engineered systems
Solid waste disposal, E-area
Intermediate-level and low-activity waste vaults
Components-in-grout trench disposal
Tank closures, F- and H-area
Saltstone disposal facility, Z-area
Decontamination and Decommissioning (D&D)
reactor buildings
canyons
other building slabs and below-grade infrastructure
4
SRNL-MS-2010-00070
Intermediate Level Vault (ILV)
5
SRNL-MS-2010-00070
Intermediate Level Vault (ILV)
6
SRNL-MS-2010-00070
Low Activity Waste (LAW) Vault
7
SRNL-MS-2010-00070
Low Activity Waste (LAW) Vault
8
SRNL-MS-2010-00070
Components-In-Grout (CIG) trench disposal
9
SRNL-MS-2010-00070
Components-In-Grout (CIG) trench disposal
10
SRNL-MS-2010-00070
Tank closures
11
SRNL-MS-2010-00070
Tank closures
12
SRNL-MS-2010-00070
Saltstone Disposal Facility
Groutplant
Vault 1
Vault 4
Future disposal
cells
Future disposal
cells
Groutplant
Vault 1
Vault 4
Future disposal
cells
Future disposal
cells
13
SRNL-MS-2010-00070
Saltstone Disposal Facility
14
SRNL-MS-2010-00070
Decontamination and Decommissioning (D&D)
Disassembly Basin
Dashed red line indicates the cross section location
Reactor Building
15
SRNL-MS-2010-00070
Decontamination and Decommissioning (D&D)
Alternative 2A:• reactor vessel remains (capped with concrete)• below-grade portions of building grouted• above-grade portion of disassembly basin removed
Assembly Area
Purification Area
Process Area
Disassembly Basin
- Existing Building Concrete (Walls, Floors, Ceiling and Roof)
- Grout
- Above Grade, Concrete Cover Over Disassembly Basin
- Ground
- Reactor Vessel
0’
-20’
-40’
0’
-20’
-40’
-51’ -51’
Water Table
NOT TO SCALE
- Concrete
Actuator Tower
ReactorVessel
- New Roofs
16
SRNL-MS-2010-00070
Key failure / degradation modes
Saltstone
sulfate attack and rebar corrosion for concrete
oxidation for HDPE liners
differential settlement / seismic events for structure
Waste tanks
acid leaching/decalcification for saturated concrete
carbonation / rebar corrosion for unsaturated concrete
general and pitting corrosion for steel tank
17
SRNL-MS-2010-00070
Key failure / degradation modes (continued)
E-area vaults
structural cracking due to seismic events and cover system overburden
rebar corrosion for concrete
Components-In-Grout trenches
grout credited for 300 years
D&D
varies
18
SRNL-MS-2010-00070
Modeling philosophy
Under, but close to,performance objective
simple is often sufficient
tailor waste stream to disposal options
no containment
components-in-grout
vaults
avoid unnecessary decontamination, cleaning, pre-treatment, and over-engineered closures → Graded approach
19
SRNL-MS-2010-00070
Modeling philosophy (continued)
Reasonable assurance
historically
deterministic analysis with conservative assumptions
sensitivity analysis
increasingly
probabilistic analysis
using abstracted models
typically a hybrid blend
deterministic results from higher fidelity model
probabilistic results from simpler, abstracted, model
Compliance?
SimplifiedTime
Mean
5%
95%
limit
Abstraction
Benchmark
Dose vs. Limit
Time (yr)
Detailed
HYBRIDHYBRID
Compliance?
SimplifiedTime
Mean
5%
95%
limit
Abstraction
Benchmark
Compliance?Compliance?
SimplifiedTime
Mean
5%
95%
limit
Abstraction
Benchmark
SimplifiedTime
Mean
5%
95%
limit
Abstraction
BenchmarkTime
Mean
5%
95%
limit
Abstraction
Benchmark
Abstraction
Benchmark
Dose vs. Limit
Time (yr)
Detailed
HYBRIDHYBRIDDose vs. Limit
Time (yr)
Detailed
Dose vs. Limit
Time (yr)
Detailed
HYBRIDHYBRID
20
SRNL-MS-2010-00070
Modeling philosophy (continued)
Core team approach
"core team" of stakeholders and subject matter experts
consensus on conceptual approach and anticipated outcomes
vetting of key results and interpretations
formalized for CERCLA / RCRA work
21
SRNL-MS-2010-00070
Modeling philosophy (continued)
Regulator constraints and considerations
CERCLA/RCRA
formal "core team" approach and decisions
agreement to use Groundwater Modeling System (GMS) graphical user interface and GMS supported codes by default
ready access to familiar software preferred
free, commonly used, well documented, open source
strong software pedigree preferred
QA, large user base and associated track record
22
SRNL-MS-2010-00070
Modeling practice
Typical practice is described
Vadose zone modeling only
no aquifer
Emphasis on
Performance Assessments
PORFLOW modeling
23
SRNL-MS-2010-00070
Software
HELP
cover system
PORFLOW
2D deterministic
GoldSim
1D probabilistic
The Geochemist's Workbench
pH and redox transitions, solubilities
STADIUM
24
SRNL-MS-2010-00070
Computational demands
Thousands of runs required
geometries
cases (best-estimate, sensitivity, . . .)
flow periods
transport species
re-runs (learning, do-overs, updated information, . . .)
Need turnaround in one to a few days
2D vadose zone with (only) several thousand nodes
temporal variability defined by a sequence of steady-stateflow fields
25
SRNL-MS-2010-00070
Computational resources
Windows (~50%)
typically multi-core personal desktops
interactive
Linux (~50%)
typically SRNL clusters
500 cores (subset available)
batch
Licenses
13 network for PORFLOW, 1+ node-locked for GoldSim
26
SRNL-MS-2010-00070
Workflow
Auxiliary supporting calculations
pH/Eh transitions
material degradation
Pseudo-transient vadose zone flow
sequence of steady-state flow solutions
single-phase, single-component
single-domain, porous-medium
Transient vadose zone transport
abbreviated radionuclide chains
linear sorption, sometimes with solubility constraints
27
SRNL-MS-2010-00070
Scripting for PORFLOW simulations
Manual approaches are often not practical
too many cases, materials, properties, species, etc.
do-overs in response to errors, new inputs, evolving needs
difficult to document and reproduce
Extensive use of pre- and post-processing scripts
more efficient, and provide a valuable electronic record
ad hoc development of several years by multiple analysts
often extensive linking (piping) of existing software
many languages and utilities
often unwieldy
28
SRNL-MS-2010-00070
Visualization
PORFLOW
Tecplot
Excel
Python
GoldSim
GoldSim GUI
Excel
29
SRNL-MS-2010-00070
Data management
Primary
Excel
Text files
GoldSim
Infrequent
"Capital D" Databases (e.g. Access)
30
SRNL-MS-2010-00070
Dimensionality
1D
probabilistic and deterministic analyses
2D
deterministic
axisymmetric or half-width
3D
rarely
D&D example
31
SRNL-MS-2010-00070
Grid resolution
D&D example
mtyp: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Tank example
5-10k nodes
32
SRNL-MS-2010-00070
Fast flow paths and fractures
Explicit fast-flow paths
usually a postulated condition
modeled as thicker seam of sand or gravel
one to a few in number
Smaller scale fractures
cracking is assumed form of concrete physical degradation
compute effective properties based on combined matrix andfracture flow
matrix and fracture assumed to be in equilibrium
33
SRNL-MS-2010-00070
Postulated fast flow path through Type IV tank
34
SRNL-MS-2010-00070
Hydraulic property variations (cementitious materials)
Frequently,
Instantaneous failure to a soil surrogate, or
Gradual increase in saturated conductivity, retainingcharacteristic curves (water retention, relative perm.)
Postulated or assumed degradation / failure
Increasingly,
Modified saturated conductivity and characteristic curves to reflect combined matrix and fracture flow
Degradation based on an auxiliary, mechanistic, analysis(e.g. STADIUM for sulfate attack)
35
SRNL-MS-2010-00070
Combined matrix and fracture flow example
B b
matrix flow fracture flow
36
SRNL-MS-2010-00070
Effective unsaturated K based on Orr and Tuller (2000)
1.E-12
1.E-11
1.E-10
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
0.1 1 10 100 1000 10000 100000 1000000
Suction Head (cm)
Un
satu
rate
d C
on
du
cti
vit
y (
cm
/s)
Intact concrete
Ksat correction only
Cracked concrete
Saturated flow
Thick film flow
Thin film flow
Active fracture
Inactive fracture
Transition zone
Matrix flow
37
SRNL-MS-2010-00070
Sulfate attack example
degradedconcrete
intactconcrete
ettringitefront
saltstone(high
sulfateconc.)
soil (low
sulfateconc.)
degradedconcrete
intactconcrete
ettringitefront
saltstone(high
sulfateconc.)
soil (low
sulfateconc.)
38
SRNL-MS-2010-00070
Vault 2 Ettringite Front
0
5
10
15
20
25
30
35
0 10000 20000 30000 40000 50000 60000 70000 80000 90000 100000
Elapsed Time (yr)
Po
sit
ion
(cm
)
Vault 2 wall
Vault 2 floor
Vault 2 roof
Ettringite formation based on STADIUM simulations
39
SRNL-MS-2010-00070
Vault 2 Effective Properties
1.0E-11
1.0E-10
1.0E-09
1.0E-08
1.0E-07
1.0E-06
1.0E-05
1.0E-04
0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
Elapsed Time (yr)
Eff
ecti
ve C
on
du
cti
vit
y (
cm
/s)
Vault 2 wall
Vault 2 floor
Vault 2 roofcompliance period
Vault 2 Effective Properties
1.0E-11
1.0E-10
1.0E-09
1.0E-08
1.0E-07
1.0E-06
1.0E-05
1.0E-04
0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
Elapsed Time (yr)
Eff
ecti
ve C
on
du
cti
vit
y (
cm
/s)
Vault 2 wall
Vault 2 floor
Vault 2 roofcompliance period
Effective saturated conductivity
40
SRNL-MS-2010-00070
Chemical property variations
Chemical equilibrium analysis
The Geochemist's Workbench
pH, Eh transitions defined in terms of pore volumes
Timing defined by flow simulation
Kd and solubility defined by pH / Eh regime, e.g.
Region II versus III in parlance of Bradbury and Sarott (1995)
Oxidizing versus reducing conditions
literature and/or SRNL experiments
41
SRNL-MS-2010-00070
Saltstone or Disposal Unit
Material
Infiltrate
Eluate = Infiltrate equilibrated with
cementitious material
Volume of Infiltrate = Volume of Eluate for Each Step
Saltstone or Disposal Unit
Material
Infiltrate
Eluate = Infiltrate equilibrated with
cementitious material
Volume of Infiltrate = Volume of Eluate for Each Step
Saltstone example
8
8.5
9
9.5
10
10.5
11
11.5
12
0 500 1000 1500 2000 2500 3000 3500 4000
Fluid Reacted (Pore Volumes)
pH
Case 1Case 2
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 500 1000 1500 2000 2500 3000 3500 4000
Fluid Reacted (pore volumes)
Eh
(m
V)
Case 1
Case 2
Case 3
pH Eh
42
SRNL-MS-2010-00070
SRS modeling interests
Multiphase, multicomponent, reactive transport
predict evolution of bulk chemistry
pH, Eh (x, t) + empirical Kd (pH, Eh) → Kd (x, t)
oxidation / carbonation of unsaturated concretes and grouts
especially for Tc-99 in Saltstone slag concretes and grouts
gas-liquid partitioning and transport in air pathway analysis
Probabilistic "full-physics" analyses
take advantage of evolving hardware
43
SRNL-MS-2010-00070
SRS modeling interests (continued)
Fractured-media approaches for cracked materials
dual-porosity? explicit fracture networks?
Open-source flow and transport code
PORFLOW replacement?
eliminate licensing limitations and "black-box" mysteries
expanded accessibility and user base
Mechanistic material degradation predictions
damage mechanics and associated properties
corrosion